How to Write and Defend a Doctoral Dissertation Successfully

What does it really take to write a doctoral dissertation that can be successfully defended?

• Perhaps native intelligence?
  
• Perhaps creativity?
  
• Perhaps the ability to get along with members of your PhD committee?
  
• Perhaps focus?
  
• Perhaps discipline?

Any of these traits can certainly make a significant contribution to success. None however account for the real reason why some PhD students succeed and so many fail to complete their PhD programs. What is the answer to this most important question?

The secret (unknown to most PhD students) is revealed in this lecture. It is one of the most important lectures in the entire course.

You have decided to earn a PhD. You are ready to devote your time and energy to revealing new discoveries to the world of science. As you immerse yourself in the process of learning research skills, you are probably overwhelmed at how much time and trouble doing a single study takes.

With this realization, you may be thinking that if you are going to have to spend 1-2 years doing a dissertation, it might as well be a study that makes a significant contribution to the field – a study that will make a name for yourself. What does this mean?

It means for most people that you carve out a very large research question that will take more than 2 years to complete. How about 6-8 years?

I made one helpful suggestion to every one of my PhD students that made it possible for them to complete their programs. Here is my speech.

“You have developed some interesting – might I say intriguing ideas here. How long do you suppose it will take to finish your dissertation?

Here is the typical answer.

“Probably a year – maybe a year and a half”

“If this were my program of research, it would take me 8 years. Did you plan on hanging around that long as a PhD student?”

Every student I ever worked with took on too many questions. They do a great job of laying out a 5-10 year program of research.

As I see it – you have 3 studies here. Why not choose one for your dissertation and set aside the other two for development later?

I am discussing this from my 20 years of experience as a faculty member. This is precisely also what I did when formulating my own dissertation as a PhD student at Michigan State University. This is what most people do from my experience.

If you are determined to finish your PhD program, you have to focus on one and only one question to ask. It sounds simple. Believe me – it is not simple. It only looks simple after you are finished doing the work.

Many PhD students have full time jobs who find it difficult to find the time to work on their dissertations. It is surprisingly easy for years to pass before students have succeeded in developing a proposal for a study, much finding the time to working on it.

Students who leave a resident status at a university are significantly less likely to complete their dissertations. I always advised my students to never get a job outside the university until they have successfully defended their dissertations. When you are out of sight, you committee members will conclude that you are not serious about doing the work and offer very little support and guidance.

This is a story of just such a person - a man who is competent, well meaning and good intentions who found himself up against the 5 years time limit to complete his PhD dissertation. If he did not defend his dissertation before the time limit expired he would be required to take and pass all of his course work and take his oral exams over again so the the clock could be reset. Of course, he had no intention of starting his program all over. He obviously needed to submit and defend a PhD dissertation before the time limit expired.

Sadly for the student who did not give himself time to write a defensible dissertation, he failed his PhD dissertation defense. Success means that you submit your work, hear the criticisms, go back to the drawing board, make the necessary changes (whatever they might be) and submit the work for another review. You will be lucky if you only have to pass through one round of revisions. If you have not engaged this process (which again is the same process used in reviewing journal articles for publication), your changes of a successful defense are slim indeed.

The point of this story is straightforward: Get started on your study immediately after passing your exams.

• Spend time on it every week.
  
• Solicit feedback from committee members throughout the process.

Give yourself plenty of time to make changes to your study so that you do not run up against the time limit set by your program to complete and defend your dissertation. Writing a PhD dissertation that can be successfully defended is a process that takes time, patience and a huge dose of persistence. If you do not like to hear criticisms, you might as well give up the idea of earning a PhD degree..

How should you structure a PhD dissertation proposal? This is not a creative endeavor! Structure the proposal the same way the final dissertation must be organized and structured.

Step back before you get too committed to a dissertation topic. Why is the study you would like to propose to your committee important? You may not get to first base if the finding you are likely to derive is obvious to everyone. If the outcome is obvious, it is not worth the trouble to bother doing the study in the first place.

Choosing a dissertation topic is challenging. Why? All the pieces of the puzzle have to fit together. You may have found or developed an awesome theory but there is no easy way to test your theory. You may have access to an incredible data set which includes a wealth of information for a very large number of subjects. Great. But, how can these data be used to test a theory and/or specific hypotheses? Linking theory and data can be a formidable challenge indeed.

There is a more straightforward approach which may save time and keep you from pulling you hair out. It is certainly not an original approach, but piggy backing onto a well formulated and designed study that has been already published may be an expedient route to completing your PhD dissertation and graduating with your PhD degree. You will not be copying what another researcher has done. It is all about building on a body of research using a different context and (usually) a different data set. You are learning how to do research by following what another experienced researcher has already done.

There is a - might I use the term - rigid structure to a PhD dissertation. Once you complete and defend a dissertation, you have demonstrated that you know how to ask an interesting question, formulate an expectation or hypothesis of what you expect to find, develop a method for testing out your hypothesis and analyze the results. Your understanding of the scientific process is demonstrated by writing to the standard dissertation "template."

1. Chapter One: Introduction
2. Chapter Two: Review of the Literature
3. Chapter Three: Theory and Hypotheses
4. Chapter Four: Methods Used to Test the Hypotheses
5. Chapter Five: Results
6. Chapter Six: Discussion and Conclusions

Your specific program may required a slightly different method of organizing the content in each chapter, but the content that is expected will not differ.

In my experience, the introduction to your PhD dissertation is one of the most challenging chapters to write. I suggest you hold off writing it until you have completed the analysis and have a clear idea about what you have found.

Are your results of your PhD dissertation divergent from other studies? If so, why? Do you confirm what other studies have shown? If so, what is your contribution to the literature? Have you explained why some studies show positive results and others do not?

In other words, how does your study fit in the context of the existent literature? You will know how to "pitch" your contribution in your introductory chapter when you have a solid grasp of the existent literature and a clear understanding of your own study findings.

The second chapter of your PhD dissertation can be challenging to formulate and write. It can be messy business when you examine the variety in how studies are done and what they find. Most students describe each study that have found that has some bearing on their study taken one by one. When the reader of the chapter gets to the end of this one by one, study by study listing of studies, no coherent understanding emerges for what is known or not known. What is the bottom line? The reader does not have a clue.

• Did all studies find the same result? If so, why are you doing your study?
  
• Is there wide variation in the results that were found? Perhaps some studies show large positive effects. Others find large negative effects. Is there any plausible explanation for why the variation is so large?

Take one specific example by way of explanation. Do the studies using farmers as units of analysis show meager results while the studies using fast food workers have large effects? When you examine the differences, do you have any good guesses as to what might explain this difference? Is it because you would expect results would be different depending on whether subjects were farmers or fast food workers? If so, does the design of your study provide the opportunity to test out your suspicion?

In summary, what is the big picture when you examine the full range of studies that are relevant to your PhD dissertation topic.

• How does the design of your study address the limitations of the existing literature?
• How are you building on the existing literature?
• Have no studies been done on your population?
• Have many studies been done, but they find different outcomes?
• Do studies show divergent results?

All things said, literature reviews are the chapter in your PhD dissertation where you make a compelling case for why your study will make a contribution to the literature (regardless of what results you find). You clearly state what is known from studies that have been done and what your contribution to the existing literature will be. You can make these claims in the literature review chapter before you have any results to report or discuss.

What is a theory anyway? A theory introduces an explanation for how the world (or people) hang together in the way that they do. Take a simple example: Are people who have been abused as children more likely - when an adult - to abuse their own children when compared to adults who did not grow up in abusive households? You could clearly choose a large number of variables and formulate a series of specific hypotheses to test this proposition.

So good so far. Simply to assert a relationship needs something more than stating it might be true. Why might this be true? It is not about "proving" it is true. That is what the test of your hypotheses does. Rather, it is all about providing a logical rationale for why whatever relationship you hypothesize might be true.

The way to develop a rationale for any hypothesis is to keep asking the magical question "why." Why would this necessarily be true (or not true for that matter). Once a sensible explanation has been formulated, you have a rationale for your hypothesis which you are now in a position to test.

Hint: Do not include in your PhD dissertation proposal the "promise" you will figure out a theory and/or hypotheses to test once you get down to the business of doing the dissertation. Develop and propose a theory and/or hypotheses before you go to the trouble of testing the hypotheses.Include the theoretical explanation in your PhD dissertation proposal. You do not want your hard work of collecting and analyzing your data to be rejected because it was not theoretically grounded.

Get your doctoral dissertation committee on board with your proposal for a theory and hypothesis (or hypotheses) at the "get go" before any data are crunched. Most committee chairs will demand that your proposal include the theory you propose to test. Do not count on it however.

Anyone can crunch numbers. Statistical packages are so sophisticated these days that it takes little in the way of expertise to generate results. Results in themselves are not a sufficient demonstration of research expertise which is what a PhD degree is all about. The degree certifies that you are qualified to conduct scientific research which begs a theoretical foundation.

What needs to be included in the methods section of your dissertation? Another researcher should be able to use the description in your methods section to conduct the same study you conducted (perhaps using a different sample or drawing a sample from a different population). Once another research can replicate your study you have provided a full and completion description of precisely how you went about defining your population, drawing your sample, defining variables and conducting the analysis. Use this criteria to evaluate whether you have covered all the issues.

Pattern the discussion of results in the results chapter of your PhD Dissertation (Chapter 5) to mirror same sequence found in the Theory chapter (Chapter 3) where you formulated your hypothesis of hypotheses. The template is:

Recall that I formulated the following expectation (remind the reader what the expectation was). Results confirm (or dis-confirm) the expectation. Then provide the details on the actual finding. Remember - some expectations are positive, some negative and some expectations formulate no specific direction - an expectation known as the null hypothesis.

Here is where you present the bottom line of what you discovered. Link this discussion to the reason you developed for doing the study in the first place What you would have already presented is a clear statement about how your results will add to the existent literature. OK - cut to the core here. What did you add to the literature? What is the take away for the reader?

Many researchers - and for my money way, way too many researchers - conclude their study by trashing it. They state all of the limitations of their study. They tell the reader in essence that they have just wasted their time reading their study. Do you really want to trash your study that you have worked so diligently to complete?

As I have said time and time again, every study has flaws. Yea, your study has flaws. So what? Many researchers hold near and dear to their heart the horrible belief that the only high quality study is a study with a control group. This is a silly idea. For some research questions, a study with a control group is the worst possible design.

Of course, your committee members will probably demand that you include a discussion of all the limitations of your study. In this case you have no choice but to do just that. When you make your study findings public through publications or presentations or newsletters, I beg you to write and speak positively about your findings.

What criteria should you consider when selecting the chair of your PhD committee? Take a close look at the publications of the candidates. Are they well published? Do they have expertise in doing the type of study you propose to do in your own PhD dissertation? Will they provide timely feedback on your proposal and your dissertation? If the candidate meets all of these criteria, they will be an excellent choice.

What qualification should you look for in the chair of your PhD dissertation committee? Sometimes, only junior faculty members have the expertise to assist with the technical nature of your study. Such persons should clearly be committee members. It is best in my opinion to select a chair of your PhD committee who has "tenure" in academic institutions that award this status. If the academic institution you are earning your doctoral degree from does not rank faculty by tenure or professorship status (full processors outrank assistant professors) then choose the faculty member who can provide the greatest technical support to your project regardless of their rank or status. You really cannot learn how to do research without mentoring from someone who knows the "game."

Faculty members in academic institutions thrive on criticism. The more criticism they get on their research the happier they are. Why? If they have received a ton of criticisms it means people are interested in their research and want to see it published in the best journal possible.

Some PhD students get discouraged when they get back page after page of criticisms from the chair of their committee. "How could my research be this bad?" they ask themselves. Some get so depressed that they have to set their research aside for months on end before they can return to it.

This reaction is completely understandable. But guess what? The student who should get really, really depressed are those who get little specific or constructive feedback. The chair of the committee can see that the project has little hope of materializing into a defensible dissertation, so the only comments the students hears are vague and general - like - "I did not quite understand what hypothesis you propose to test and why it is interesting?" When you get these types of vague responses that give you little clue as to how to fix the problem it is time to go back to the drawing board and overhaul the project.

You have worked very hard on your dissertation project and have at long last submitted the first three chapters for review, comment and feedback by your committee chair.

• One week passes. No response is received. This is OK with you actually. You need a little break from working on your dissertation anyway.
  
• One month passes. Still no response is received. This is still OK with you actually. You need a little longer break from working on your dissertation.
  
• Two months pass and still no response from your chair. Gee Wiz you think. Well, my chair is very busy I will be patient and wait to hear back from them.

And so it goes, Month after month pass by and you still do not get any feedback. What is up doc?

It would seem at first glance to be a good idea if a member of your PhD committee has not taken the time to read and evaluate your dissertation before you defend it. If they have not read your dissertation, how could they have any questions for you? How would they be in a position to criticize your study?

I strongly advise that you dump this thinking. I can assure you that any committee member of your dissertation does not have to read one sentence from your dissertation to criticize what you have done. That is what academics are all about! They are great at criticizing research.

You will be in a much stronger position if you get all of your committee members to read your dissertation (probably not thoroughly as your chair will do, but at least look over your project). Request that they give you comments and criticisms before you defend your dissertation. You may well need to make revisions to your study (or even re-do some of what you have already done) in light of the issues that are raised. It is far more expedient to fix the problems before the defense than having to defend a second time around.

Of course, you have already spent months if not years on your study. You are more than ready to be done with it. If your chair has approved your study for defense, it is tempting to go for it.

I recommend that you pause, take a deep breath and seek out comments from all members of your committee before you defend. No. Of course you do not want to edit, revise or change your study in any way. But this is what the review process is all about.

You may have a fantastic study. You may be very proud of it. You may think that criticisms you have heard from a committee member are baloney. Do not miss the bottom line. You have to get the approval of your committee members before you have passed through this hurdle.

The process of writing a PhD dissertation and defending it successfully is no different from how a journal article is reviewed and considered for publication. Many revisions often have to be made (and sometimes additional data or evidence gathered) before a panel of journal reviewers (or your PhD committee) will say "good enough." If you are not willing to make revisions or changes to your work, you might as well give up working on getting your PhD degree right now!

PhD students are always excited and eager to defend their PhD dissertation once their dissertation chair tells them it is ready to defend. Here is the common reaction:

Finally! I have made so many changes and revisions to my study. I am actually getting a little sick of my PhD dissertation. Looks like the light is finally green to move on to the final step of approval. I am scheduling a defense ASAP."

Might I suggest that you put the plan to defend your dissertation on pause? I have another approach that increases the odds of a successful defense significantly. You can go for it if your chair has given you the OK to be sure. But consider this. Do you want to walk into your defense and hear criticisms of your study that you never knew were problems? This is called being blind sided. Would it not be a smarter choice to have time to prepare responses to any and all criticisms? You can make changes and improvements that are required by your committee before you defend. You are going to have to make the revisions anyway!

You will likely still have to make some changes after the defense before your committee chair signs off on your dissertation. The intention is to be in a position to have reasoned responses to the criticisms during your defense that make it clear to the committee that you now know how to conduct a scientific study. You do not want to be blind sided. A PhD degree signifies you are qualified to conduct research that is publishable and that makes a contribution to your field, This is why you are required to defend your study.

Many PhD students initially identify a rich body of studies that pertain to their PhD dissertation topic. The studies fall into many different categories including quantitative field studies, laboratory studies, reviews of the literature, case studies and others. Because the body of work is so diverse, the temptation is to narrow down the number of studies that are included in the literature review chapter of the dissertation.

For example, let us say you have located 50 studies that are relevant to your topic. Only six used control groups. It is so much work to review 50 studies so you think to yourself: It would be so much easier to simply include in the literature review only the "high quality" studies that used control groups. This would leave only 6 studies to include in the review rather than 50 (ugh)! The temptation is great because reviewing six studies takes so much less effort and time than going to the trouble of summarizing and making sense out of 50 studies.

My strong recommendation for you is to include all 50 studies in the review (in this particular example). Consider the full range of variation in the studies that have been done.

There will likely be academicians who will try to convince you otherwise. They believe that a study must meet certain conditions to be a "high quality" study. Every study is flawed in one way or another - even studies that have control groups.

To summarize, include all of the studies in your review. You will be thankful you did when you realize what you can see and discover when you have the full range of outcomes to consider.

You are stumped. Why? You have identified a great topic for your dissertation but there are so many studies that have already been published on your topic. Even worse is your observation that the variation in study outcomes is huge. Some studies find large positive effects. Other studies seems to report negative effects. How can this be really?

It all seems like a horrible mess. It will take months and months if not years to just make sense out of the existent literature.Finishing the second chapter of your dissertation seems to be an impossible task and this is only one of your chapters to complete!

If this has happened to you, I have a recommendation for you to consider. Propose a PhD dissertation that conducts a systematic, quantitative analysis of all studies. The unit of analysis of your PhD dissertation in this case is studies that have been done on the topic (including published and unpublished as well as other dissertations). In other words, instead of the unit of analysis being a person or an organization or a country,your unit of analysis can the a study. How cool is that?

The technology for such a study is known as a meta-analysis (or a cumulative analysis of other study findings). This happened to be one of my own areas of expertise so I was able to help my own students with such studies. The technology for doing a meta-analysis study is now well developed, with many excellent software programs taht can be used to convert study findings to a common metric. Meta-analyses are quantitative studies, so if you prefer the qualitative approach, you will not want to pursue this option.

So, it is now time to decide whether you will choose to do a quantitative PhD dissertation or a qualitative PhD dissertation. Do you want to crunch numbers or interview people for a case study or set of case studies? Many students have the impression that quantitative studies are more difficult to do. I have a surprise for you!

In my experience, case study or qualitative studies are far more challenging and difficult to conduct and complete than quantitative studies. If you have decided to pursue a qualitative dissertation, be sure to detail out in your PhD dissertation proposal precisely how you propose to proceed, what information you will gather, how you will gather it and most importantly, how you plan on discussing the results that you get. Without a clear proposal of how you will proceed, you may be working on finishing your qualitative case study PhD dissertation for years on end. Why? There is no clear stopping point. You committee may keep telling you to gather more information or analyze it in a different way. Anchor everything down at the outset.

You obviously want to get to work on writing the dissertation itself and will get frustrated with having to rework your proposal time and time again. The hard work of research is in developing the proposal for a study. Once you have the proposal in good shape, doing the actual work will be easy in comparison to developing the proposal.

One problem that can be a "show stopper" for a dissertation study is what is known as "restriction in range". When the variation in the dependent variable is restricted, you will not be in a position to test any hypotheses. Be sure that the data you collect for your dependent variable - the focus of what you are interested in explaining - represents the highs as well as the lows of possibilities. To put it another way - the analysis should consider a sample of the full spectrum of possibilities.

What exactly is mean by "survey research"? This approach to collecting data involves developing a set of questions on a questionnaire that is answered by your study sample. The questionnaire can be completed in paper form or completed electronically.

Although you may be able to use instruments that have already been developed and validated by another researcher, you often must develop a questionnaire instrument yourself. Why? You will seldom find a questionnaire that includes the variables that you will need to test the hypotheses you need to test. The questionnaire must include questions that provide the data you will need - both dependent and independent - which can be used to test your hypotheses.

What constitutes a random sample? Each member of the population must have an equal probability of selection. If you select, say, one out of every 10 members that are listed in your population data set, you will not be drawing a random sample.

It is pretty easy to draw a random sample from a data set using statistical packages. The problem with random sampling often turns on drawing a study sample "by hand" - the old fashioned way. Make sure to use a random number generator to select members of your study sample if you are drawing your sample "by hand" and do not have the advantage of accessing an population data set that is digitally coded and selecting the sample with a standard statistical package like SPSS or SAS. .

Many PhD students and doctoral students are on the lookout for the perfect measure of their dependent variable. Perfect measures do not exist. Every variable has a degree of error in measurement. Some measures have very low reliability.

Instead of being stuck on using a single variable to measure for your dependent variable, identify several variables that can be used to create an index. The index becomes how you measure your dependent variable. Indexes are formed by adding up the standardized measures of the variables you have chosen to include in the index.

How do you decide which variables to use in the index? Examine the inter-item correlations in your data set. Variables that measure the same thing will be highly (but not perfectly) correlated.The other key indicator to use is face value. Is it logical to that the variables are measuring the same thing? Of course, any variables used to as predictor (or independent variables) can also be indexed.

To summarize: The problem of measurement error (or problems with the low reliability of a variable) are addressed in part by constructing an index which includes more than one variable.

Sometimes it is indeed difficult to sort out the reasons you are having trouble completing and defending your dissertation.  I am happy to offer consultations to help out with any aspect of your study - whether it is content, working with your committee and/or chair or analyzing and presenting your results.

Feel free to call me so we can sort out together the issues and how I might be able to provide the support you need. The goal here is to finish and successfully defend!

Robert Rodgers PhD
Olympia Washington USA
360-789-1658